Planned vignettes

[kaitejohnson]

Goal
Outline desired vignettes for baselinenowcast

Proposed vignettes

1. Getting Started (baselinenowcast.Rmd): shows a straightforward example of estimating a delay distribution from a single strata and using it to nowcast the same strata. See Set up initial problem statement in the vignette #4 for discussion of what dataset to use for this and how to interface with epinowcast
2. Extensions to multiple strata and different temporal granularities (extensions.Rmd): expands on the initial vignette, demonstrating how to "borrow" a delay estimate from one reporting triangle and apply it to another, and gives an example from another dataset with a different temporal granularity in either reference dates or report dates.
3. How it works (nowcast_definition.Rmd): walks through the mathematical details of the "model", how uncertainty is estimated, and how days with no observations are handled
4. How does baseline nowcast compare to other methods (baseline_comparison.Rmd): Walk through an analysis that produces nowcast for a range of scenarios using both baselinenowcast, epinowcast, and potentially other methods. Will fully flesh out what sensitivity analyses we want to do (e.g. varying data sparsity of the strata, impact of estimating on data with weekday event or reporting effects) and what data we want to use in a separate issue.

Alternatives

• Anything else we should consider? The only other thing I was thinking of was a separate vignette to show how someone who get their data from line-list at the visit level into the input they need it for baseline(), potentially both with and without relying on enw_preprocess().

Additional context
#1 (comment)